Return-Path: X-Original-To: apmail-hbase-issues-archive@www.apache.org Delivered-To: apmail-hbase-issues-archive@www.apache.org Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by minotaur.apache.org (Postfix) with SMTP id E5F7111E82 for ; Sun, 24 Aug 2014 16:29:38 +0000 (UTC) Received: (qmail 46575 invoked by uid 500); 24 Aug 2014 16:29:38 -0000 Delivered-To: apmail-hbase-issues-archive@hbase.apache.org Received: (qmail 46529 invoked by uid 500); 24 Aug 2014 16:29:38 -0000 Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Delivered-To: mailing list issues@hbase.apache.org Received: (qmail 46518 invoked by uid 99); 24 Aug 2014 16:29:38 -0000 Received: from arcas.apache.org (HELO arcas.apache.org) (140.211.11.28) by apache.org (qpsmtpd/0.29) with ESMTP; Sun, 24 Aug 2014 16:29:38 +0000 Date: Sun, 24 Aug 2014 16:29:38 +0000 (UTC) From: "Andrew Purtell (JIRA)" To: issues@hbase.apache.org Message-ID: In-Reply-To: References: Subject: [jira] [Commented] (HBASE-11813) CellScanner#advance may infinitely recurse MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-JIRA-FingerPrint: 30527f35849b9dde25b450d4833f0394 [ https://issues.apache.org/jira/browse/HBASE-11813?page=3Dcom.atlassia= n.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D141= 08443#comment-14108443 ]=20 Andrew Purtell commented on HBASE-11813: ---------------------------------------- Thanks for reporting back [~Schabby]. Please let us know if this is still l= ooking good after a day or so.=20 > CellScanner#advance may infinitely recurse > ------------------------------------------ > > Key: HBASE-11813 > URL: https://issues.apache.org/jira/browse/HBASE-11813 > Project: HBase > Issue Type: Bug > Reporter: Andrew Purtell > Assignee: stack > Priority: Blocker > Fix For: 0.99.0, 2.0.0, 0.98.6 > > Attachments: 11813.098.txt, 11813.098.txt, 11813.master.txt > > > On user@hbase, johannes.schaback@visual-meta.com reported: > {quote} > we face a serious issue with our HBase production cluster for two days no= w. Every couple minutes, a random RegionServer gets stuck and does not proc= ess any requests. In addition this causes the other RegionServers to freeze= within a minute which brings down the entire cluster. Stopping the affecte= d RegionServer unblocks the cluster and everything comes back to normal. > {quote} > Subsequent troubleshooting reveals that RPC is getting stuck because we a= re losing RPC handlers. In the .out files we have this: > {noformat} > Exception in thread "defaultRpcServer.handler=3D5,queue=3D2,port=3D60020" > java.lang.StackOverflowError > at org.apache.hadoop.hbase.CellUtil$1.advance(CellUtil.java:210) > at org.apache.hadoop.hbase.CellUtil$1.advance(CellUtil.java:210) > at org.apache.hadoop.hbase.CellUtil$1.advance(CellUtil.java:210) > at org.apache.hadoop.hbase.CellUtil$1.advance(CellUtil.java:210) > [...] > Exception in thread "defaultRpcServer.handler=3D5,queue=3D2,port=3D60020" > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D18,queue=3D0,port=3D60020= " > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D23,queue=3D2,port=3D60020= " > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D24,queue=3D0,port=3D60020= " > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D2,queue=3D2,port=3D60020" > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D11,queue=3D2,port=3D60020= " > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D25,queue=3D1,port=3D60020= " > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D20,queue=3D2,port=3D60020= " > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D19,queue=3D1,port=3D60020= " > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D15,queue=3D0,port=3D60020= " > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D1,queue=3D1,port=3D60020" > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D7,queue=3D1,port=3D60020" > java.lang.StackOverflowError > Exception in thread "defaultRpcServer.handler=3D4,queue=3D1,port=3D60020" > java.lang.StackOverflowError=E2=80=8B > {noformat} > That is the anonymous CellScanner instance we create from CellUtil#create= CellScanner: > {code} > =E2=80=8B return new CellScanner() { > private final Iterator iterator =3D cellSc= annerables.iterator(); > private CellScanner cellScanner =3D null; > @Override > public Cell current() { > return this.cellScanner !=3D null? this.cellScanner.current(): nu= ll; > } > @Override > public boolean advance() throws IOException { > if (this.cellScanner =3D=3D null) { > if (!this.iterator.hasNext()) return false; > this.cellScanner =3D this.iterator.next().cellScanner(); > } > if (this.cellScanner.advance()) return true; > this.cellScanner =3D null; > ---> return advance(); > } > }; > {code} > That final return statement is the immediate problem. > We should also fix this so the RegionServer aborts if it loses a handler = to an Error.=20 -- This message was sent by Atlassian JIRA (v6.2#6252)